IR Transmitter for smartphone

My understanding problem is as follows : on figure 1 for the signal 2 (noted: S2) . If I understand correctly, the LED cannot reach S2 signal because polarity. Carrier Frequency is 38 KHz let 38 000 Hz (carrier freq) / 2 (channels L+R)) = 19 000 Hz / channel, then i can reach only 19 KHz ? But the understanding problem is my TV is successfully controled by this frequency, but why ? 19000Hz is far too far from the frequency carrier. 38 000 ~+20% or ~-20% from IR receiver general tolerance. I start in electronics it is possible that I miss or did not understand a certain part.If someone has an idea and time to grant me, thank you.

My 'problem' is why (not how) with only 1 LED and a stereo file i can control my TV ?

• I have tried the same like you before (Galaxy S3 and other phones) but can't get it to work. Bought one in china for about 0,85 and doesn't work either. I think it is a hoax, you need to make a powered IR-blaster, search on google, hackaday or instructables. – Codebeat Jan 29 '17 at 19:49
• Anaway, two leds are required to reach 38Khz. 38Khz is ultrasound and your audio socket is unable to produce this. Take a look at your amplitude, both sides are used to reach 38Khz (19+19). You need audicity to modify the wave for this purposal. – Codebeat Jan 29 '17 at 19:54
• Thank's but for me it's work in the two case. I don't understand why with only one IR LED it's work while carrier frequency is 38KHz. With 1 LED frequency is 19 000Hz and working distance is centimeters while 2 LEDs frequency reach is 38 000Hz and working distance is meters.... – Ephemeral Jan 29 '17 at 19:55
• You need audicity to modify the wave for this purposal. Google it. – Codebeat Jan 29 '17 at 19:56
• wiki.samygo.tv/index.php5/Build_your_own_IR_transmitter – Codebeat Jan 29 '17 at 19:58

A perfect 50% square wave pulse has no even harmonics. So a perfect 19kHZ square wave will not work on a 38kHz Rx with a BPF.

The greater the asymmetry of the IR pulse, the odd harmonics reduce while even increase to an impulse which has all harmonics nearly equal to fundamental ( dropping quickly with null at wavelength equal to pulse width)

There are various methods to create 2nd harmonics but beyond the scope of this question.

The easiest method to create 2nd harmonics is to put the audio into a full-wave rectifier then a suitable driver.

Using the Falstad Java simulation of Fourier analysis you can play with the tool captured below. Going the website shown and selecting the buttons pointed by arrows.

The cursor shows the component of each harmonic that makes the waveform with extra information on the spectrum . The 1st spike is DC then the 2x, 4x, 6x etc shown in the log (dB) view above simply by full-wave rectifying one sinewave.

simulate this circuit – Schematic created using CircuitLab

• Thank you very much clarified this answer but I have trouble understanding what I'm doing, so I use audio sine waves (analog signal input) but they are interpreted as square waves by the receiver (digital output signal)? I looked for documentation on harmonics but not found a really good documentation. On rfid (nfc) I search and it seems to be if I double frequency or triple etc ... I can see the data transmitted on my software sdr. Is this what we call harmonic? – Ephemeral Jan 29 '17 at 21:11
• – Ephemeral Jan 29 '17 at 21:19
• Thank you very much, you had already answered me on the same subject a few days ago, thanks to you I discovered the site falstad and there I learn a new functionality without mentioning all the information you bring me that Are just and constructive. I do not know how to thank you. – Ephemeral Jan 29 '17 at 21:50
• Thank's for the diagram and help me to learn electronic and waves.:) – Ephemeral Jan 29 '17 at 21:54
• Ok, second harmonic (which I discover) I do not think because I have no hardware other than a led IR 940nm and a stereo jack without microphone – Ephemeral Jan 29 '17 at 22:28

This is for the old timers and also for those younger persons who are not followed smartphone hacks. This time is attempted to insert IR leds into the earphone socket to make IR remote controller possible. There's a problem. IR remote systems need about 36...39 khz pulsed led drive voltage and normal audio systems can only output about 20 kHz due the sampling frequency 44,1 kHz

This is outsmarted by pushing to left and right outputs 90 degrees phase shifted signals and by using two leds in parallel between LEFT HOT wire and RIGHT HOT wire. In the following picture is illustrated, how there still can be 38 kHz pulses available altough both audio channels output 19 kHz.

To get IR pulse from both positive and negative "Difference" pulses, two parallel, but opposite IR leds are required plus a series resistor to keep the current low enough.

I have drawn squarewaves. In reality only sinewaves are available over 10 kHz due the limited frequency band, but that does not change the basic idea. It works, if the signal is driven hard enough to get enough distortion (the difference of sinewaves is a sinewave)

The IR add-ons for the earphone socket are widely available and the remote control application software, too.

ADDENDUM: Seemingly I forgot the main question "How this can work with one diode at least at few centimeters"

In the IR receiver ICs the right pulse frequency is discriminated by bandpass filter. That filter can get enough 38 kHz signal altough the pulse is 19 kHz by the following principles:

• 2nd harmonic in this narrow L-R difference pulse is >0 altough in symmetric pulse it's =0. Thus one led between L and R can be enough for short distance
• severe ringing in smartphone's DAC output filter can create the missing pulse
• at high IR intensity the receiver can generate the 2nd harmonic

It is maybe not at all the 2nd harmonic, but you have Bidirectional IR LED. It is 2 opposite parallell leds in one case. Thislike devices are available easily. EDIT: Not easily. They are common only for light and in YouTube videos they can be fake.

Which of these reasons are the remarkable ones in practice, must be checked by measurements. All other is only quessing.

• thank's but you call about doubling frequency only ? My problem is : It's work with only one ir led and then the frequency is not the double because polarity of Right or negative channel are inverted from ir led (already used for Right or negative channel). This may be a physical or other effect (second harmonic ??? thank's Tony). I assert that with only one led this works for frequency 37.47KHz (LG European TV) and Vestel manufacturer with Mitch Altman POWER-OFF codes reproduct to sinusoidal stereo waves at sample rate 44100 Hz 16 bit PCM wav file (audacity), 1 led transmitter. – Ephemeral Jan 29 '17 at 22:12
• @abort Edited the answer for the comment – user287001 Jan 29 '17 at 22:53
• Yes indeed I did not see the very important part of your answer on the SHG ... Thank you very much I will go to learn the second harmonics that complement me unknown :) – Ephemeral Jan 29 '17 at 23:01
• youtube.com/watch?v=BkD0LOBZk4c distance in this movie is long, it is not 1st but probably your 2nd or 3th answer. – Ephemeral Jan 29 '17 at 23:16
• @aborted A new addition inserted: Maybe you have a bidirectional led – user287001 Jan 29 '17 at 23:32